This forecast is a tough one. So tough, I cloned myself just to be able to look at the mountain of data.
So much for the fun part. These things are always tough. There are always a lot of things to look at, but the priorities change each winter. This is not like forecasting for the next 10 days. Computer models are less useful for seasonal forecasts.
Breaking news and the stories that matter to your neighborhood.
Factor #1: The Pacific
The first thing we look at is what’s happening in the Pacific, especially the Tropical Pacific. Is it an El Niño or La Niña? This year, it’s neither. If anything, there are some signs that the Central Pacific part of what we look at will be warmer than the rest of that area, known as a Modoki El Niño. Research has shown that Modoki El Niños have been occurring more often since about 1990. This may be related to the world’s changing climate. And the last three Modokis have resulted in snowy and cold winters for us.
Factor #2: Trends
"The trend is your friend." That’s a much-used phrase in the financial world related to investing. But it’s also true in weather forecasting. I always check to see if there are any strong trends in statistics for the Philadelphia area. And there are.
First, let me say that I no longer look at any data before 1990, and not much before 2000. The climate has changed so much in the past 20 years that I have found many of the old “rules” are no longer valid in winter forecasting.
The map below shows average temperatures compared to normal for each October since 2012. Warm, just like this year.
And the next map is for November since 2012. What a difference! Way colder than normal in the Eastern United States. Just like this year.
So, what about December?
Another amazing correlation! A strong trend toward milder than average Decembers over the past 6 years.
The same trends don’t hold up for January or February. However, there’s a real strong trend for cold and snow in March.
Factor #3: The Sun
The sunspot cycle has been very stable and predictable. It’s not enough to significantly affect the climate. The proof of that is in the picture below. It shows how the overall activity has been decreasing since 1990, yet the global temperatures have been increasing.
But it does seem to have an influence on seasonal weather. We are at a solar minimum this year, so I wanted to see if there was a correlation of previous minimums with winter weather. And there sure is!
Previous solar mins (including a 2-3-year period around the min) have included two of our snowiest winters on record (2009-10 and 1995-96). And the trend goes back at least to 1964-65. Of the 10 winters studied, NINE of them had below normal winter temperatures! And the median (always use median rather than mean for snow) was 27”, which is 5” above average.
To get true correlations, I also want to look at the solar maximums. They should deliver above normal temperatures and below normal snowfall. And they do!
We should also see what average weather patterns look like during the solar mins. Below, are the mean December through March maps around 20,000 feet up in the atmosphere, which is the level we look at for such things.
And what did I find? A massive "positive anomaly" over Greenland (in the reddish color). Negative anomalies extended of the Eastern U.S. across the Atlantic to Europe/ This is a CLASSIC pattern known as a negative North Atlantic Oscillation, or -NAO. Winters with prevailing -NAO have strongly trended to be colder and snowier than average. You’ll never see this stat anywhere else: Low solar winters were FIVE times more likely to average -NAO than +NAO!
Factor #4: October Snow in Eurasia & Northern Hemisphere
A funny thing has been happening in much of the northern part of the world. We all know the Arctic has warmed much more quickly than the earth has over the last few years. But autumn snowfall has increased across the Northern Hemisphere and across the world in Eurasia. That trend has not occurred during winter itself.
Since 1990, the trends seem clear. More snow. The reason this is important is that research has correlated October snowfall in these areas means snow in our part of the world, months later during the winter. These high levels of October snowfall there are yet another factor that favors more snow for us here.
Factor #5: Soil Moisture
We have noticed over the years that areas of the country with the most soil moisture have an influence on future weather patterns. Specifically, the upper-air "troughs" tend to form over the moist area during the upcoming season.
This year, the soil moisture is clearly concentrated in the Eastern United States.
Factor #6: The Blob
The picture below shows current ocean temperatures (compared to "normal") all across the entire world. The most dramatic feature is in the Northeast Pacific, off the southern coast of Alaska and Western Canada. That large area of warm water can lead to patterns that bring cold air into the Eastern United States.
And here’s a close-up comparing the current "blob" with the one in 2014.
You may ask, "What was the winter of 2014-15 like?" Well, it happened to be the last winter we had with below normal temperatures. There was way above normal snow cover during October in Eurasia and the Northern Hemisphere. For us, October turned out warm, and November ended up cold. Sound familiar?
We had 27 inches of snow that winter, which was above normal, but not by much, considering how cold it was. Our snowiest month was March, with 12 inches. That winter was also near a solar maximum.
Like most seasonal forecasters, I also look at other ocean indices such as the PDO [Pacific Decadal Oscillation], and even high up in the stratosphere (the QBO [Quasi-Biennial Oscillation]. Both are in favorable positions for the colder and snowier winter that my other factors indicate.
In December, I’m predicting a mere 3 inches of snow, and well above-normal temperatures.
The pattern changes in January, with 12 inches of snow and temperatures far below normal.
February reverses again, with milder temps, but still delivers about 10 inches of snow.
In March, the temps drop and so does another 8 inches of snow.
This all adds up to slightly below normal temperatures for the winter. But in a warming climate, that is hard achieve. So, it will likely be the coldest winter since 2014-15.
The 30-35 inches of snow for Philly is about 10 inches above normal. I would say there’s about a 20 percent chance of more than 40 inches, and a rather low 20 percent chance of less than 20 inches.
Will the blob break up? Will the El Nino pattern change? There are several unknowns, as usual. But that’s what winter season forecasting is all about.